STATISTICAL METHODS AND APPLICATIONS IN HUMAN GENETICS

数学和统计方法的发展和应用 用于复杂人类特征的遗传分析。 在 特别是，主要重点是制定方法， 通过遗传学来理解复杂的遗传模式 连锁标记 限制性片段长度的出现 跨越整个基因组的多态性（RFLP）现在提供了 这样的分析能力。 具体来说，我们打算： 涉及多个基因座（例如，上位性和遗传性） 异质性模型）;确定检测 这种复杂的模型;确定建立复杂的 多个连锁标记的遗传模式。 我们还将 检查不同的发病年龄作为遗传的潜在指标 异质性及其对连锁分析的影响。 最后我们 将探讨应用连锁分析多变量 性状，并决定连锁的力量，以解释遗传 性状间相关性的来源。 所开发的方法将应用于三个主要数据 集. 首先，多变量数据，包括家族史， 特发性扭转性肌张力障碍的检查目标是 建立遗传异质性;这将具有重要意义 建议的家庭和连锁研究的后果。 二是 大样本乳腺癌家族数据（4700名先证者）将 检查以确定遗传异质性的可能性 由发病年龄和可能的其他风险因素决定; 这些结果将对乳腺癌的连锁分析产生影响。 第三，一组关于免疫反应的广泛数据， 南美印第安人肺炎球菌抗原与HLA、Gm和Km 将从多变量遗传学的角度分析家庭 以确定遗传模式的相关性。 HLA、GM的作用 将确定家族模式中的Km。 预计开发的方法将具有广泛的应用前景。 适用范围超出上述具体项目。 的 该项目的结果将具有重要的临床和公众意义。 健康的重要性，因为他们将提供一个更好的了解 疾病病因学和（可能复杂的）遗传贡献 并为治疗设计，预防， 准确的风险评估。

Development and application of mathematical and statistical methods for the genetic analysis of complex human traits is proposed. In particular, the primary focus is on developing methods for understanding complex patterns of inheritance through genetic linkage markers. The advent of restriction fragment length polymorphisms (RFLP's) spanning the entire genome now offers the power to achieve such analysis. Specifically, we intend to: elaborate models of inheritance involving multiple loci (for example, epistasis and genetic heterogeneity model); determine the power of detecting linkage for such complex models; determine the ability to establish complex patterns of inheritance with multiple linked markers. We will also examine variable age of onset as a potential indicator of genetic heterogeneity, and its impact on a linkage analysis. Finally, we will examine the application of linkage analysis for multivariate traits, and determine the power of linkage to explain genetic sources of correlation among traits. The methods that are developed will be applied to three major data sets. First, multivariate data including family history on idiopathic torsion dystonia will be examined with the goal of establishing genetic heterogeneity; this will have important ramifications for proposed family and linkage studies. Second, a large sample of family data on breast cancer (4700 probands) will be examined to determine the possibility of genetic heterogeneity as determined by age of onset and possibly other risk factors; these results will impact on linkage analyses of breast cancer. Third, an extensive set of data on immune response to twelve pneumococcal antigens and HLA, Gm and Km in South American Indian families will be analyzed from a multivariate-genetic standpoint to determine genetic patterns of correlation. The role of HLA, Gm and Km in the familial patterns will be determined. It is anticipated that the methods developed will have broad applicability beyond the specific projects described above. The results of this project will be of significant clinical and public health importance because they will offer a better understanding of disease etiology and the (possibly complex) genetic contribution to it, as well as provide avenues for treatment design, prevention, and accurate risk assessment.